In conventional computing environments implementing a hypervisor to execute a virtual machine on a host computing device, the hypervisor typically provides the virtual machine with access to hardware resources provided by the host computing device. In such an environment, a conventional hypervisor does not typically provide functionality for preventing errors that result from overuse of an allocated resource from impacting a user of a virtual machine. For example, if a resource uses more than its allocated processor time or allocated disk access, the resource may create conflicts with other resources or cause an ungraceful termination of the virtual machine or of other virtual machines sharing access to the physical disk. Conventional hypervisors may receive an indication that an error or conflict has occurred after the fact, or determine, prior to execution of a virtual machine, that an error or conflict may occur. However, conventional hypervisors do not typically provide flexible options for continuing execution of a virtual machine after the occurrence of such errors.